(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The Indo-European Cattle-Raiding Myth"

§ I CHICAGO JOURNALS 



The Indo-European Cattle-Raiding Myth 

Author(s): Bruce Lincoln 

Reviewed work(s): 

Source: History of Religions, Vol. 16, No. 1 (Aug., 1976), pp. 42-65 

Published by: The University of Chicago Press 



Stable URL: http://www.jstor.org/stable/1062296 
Accessed: 28/04/2012 16:18 



Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at 
http ://www.j stor.org/page/info/about/policies/terms .j sp 

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of 
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms 
of scholarship. For more information about JSTOR, please contact support@jstor.org. 




The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to History 
of Religions. 



STOR 



® 
http ://www.j stor.org 



Bruce Lincoln 



THE INDO-EUROPEAN 

CATTLE-RAIDING 

MYTH 



In an earlier paper entitled "The Indo-European Myth of Crea- 
tion," 1 I attempted to trace out a myth of Indo-European 
provenance that detailed the events which were believed to have 
been crucial in the establishment of the world as the Proto-Indo- 
Europeans knew it. There I argued that this myth told of two 
brothers, *Manu- "Man" (Sanskrit Manu, Avestan *Manus, 
Germanic Mannus being linguistic correspondences; Old Norse 
Odinn and Latin Romulus being structurally related) and *Yemo- 
4 'Twin" (Sanskrit Yama, Avestan Yima, Old Norse Ymir, and 
Latin Remus being linguistic matches; Germanic Tuisco being a 
semantic match; Sanskrit Mandvi and Purusa, Pahlavi Gayomart 
being structurally related). Originally, this myth told of how 
*Manu, a priest, sacrificed *Yemo, a king, together with a bovine 
animal, and then created the world from their respective bodies: 
the physical world and the three I-E social classes (sovereigns, 
warriors, and commoners) 2 coming from the body of the sacrificed 

1 History of Religions 15, no. 2 (November 1975): 121-45. See also the article of 
Jaan Puhvel, "Remus et Frater," in the same issue, which supports and extends 
the argument. 

2 These classes have been well established by the work of Georges Dumezil. 
For a convenient summary of his views, see his & Ideologic tri-partie des indo- 
europeens (Brussels: Collection Latomus, 1958); or C. Scott Littleton, The New 
Comparative Mythology : An Anthropological Assessment of the Theories of Georges 
Dumezil, 2d ed, (Berkeley: University of California Press, 1973). I do differ with 
Dumezil on the nature of the so-called Third Function and regard it as something 
of a catchall class for anyone not belonging to the upper classes. Thus, I prefer the 
term "Commoners" to any of his designations. 

42 



History of Religions 

king, while the animal and vegetable species came from the 
sacrificed bovine. 

In this myth I felt that we could see something of the mythic 
"charter" 3 which the Proto-Indo-Europeans established for them- 
selves : the differentiation of priests and kings as differing types 
of the sovereign; the priest's role as ritual specialist, the king's 
as that of complete man ; the separation of the three social classes ; 
the unique position of cattle as the intimate companion of man 
ab origine and source of all good things; and, last, the crucial im- 
portance of sacrifice in the creation and preservation of the world 
order. 

This myth of creation seemed to me to be of fundamental 
importance for I-E society. What I did not recognize while pre- 
paring that earlier article was the fact that the myth of *Manu 
and *Yemo focused almost exclusively upon the twin figures of 
sovereignty: priest and king, whom Dumezil has grouped together 
in his "first function." 4 It is thus with a sense of some excitement 
that I feel I can now point to another myth which seems to 
"complete" the cosmogony: a myth which deals primarily with 
the nature of the warrior function. 5 Moreover, the very name of the 
hero in this myth — *Trito- "Third" — is extremely suggestive of 
an old I-E mythic cycle which grouped together these three pri- 
mordial heroes, "Man," "Twin," and "Third," each of whom 
served as the mythic model for a different social group. 

Actually, the myth with which we will be dealing is one which 
has long been familiar to scholars. 6 In truth, it combines two of the 
best known mythic themes: that of slaying a serpent or monster, 
and that of stealing a neighbor's cattle. The former is almost 

3 Used in the sense proposed by Malinowski in his essay "Myth in Primitive 
Psychology," in his Magic, Science and Religion and Other Essays (Garden City, 
N.Y.: Doubleday & Co., 1948), esp. pp. 107-8. 

4 See Dumezil's Mitra-Varuna, 4th ed. (Paris: Gallimard, 1948), pp. 23-26, 
and "Le Rex et les Flamines Maiores," in La Regalita Sacra (Leiden: E. J. Brill, 
1959), pp. 407-17. While Dumezil has on occasion shown hesitation on including 
kingship in the sovereign class, as in "Religion indo-europeenne: Examen de 
quelques critiques recents," Revue de Vhistoire des religions 152 (1957): 15-16; 
the argument advanced by Emile Benveniste, "Traditions Indo-Iraniennes sur 
les classes sociales," Journal Asiatique 230 (1938): 534-35 is most convincing. 

5 On the warrior class, see especially Stig Wikander, Der arische Mannerbund 
(Lund: Gleerupska Universitet Bokhandeln, 1938); Dumezil, Destiny of the 
Warrior, trans. Alf Hiltebeitel (Chicago: University of Chicago Press, 1970). 

6 Among the earliest treatments were Rudolf von Roth, "Die Sage von Feridun 
in Indien und Iran," Zeitschrift der deutschen morgenlandischen Gesellschaft 2 
(1848): 216-30; Michel Breal's "Hercule et Cacus," in his Melanges de mythologie 
et de linguistique (Paris, 1882), pp. 1-62, first published as his dissertation in 
1850. While chronologically much later, the treatment of Leopold von Schroeder, 
Her aides und Indra (Vienna: Denkschriften der kaiser lichen Akademie der 
Wissenschaften in Wien, 1914), pp. 57-67, should be grouped with them. 

43 



Indo-European Cattle-raiding Myth 

universal in its dispersion, being seen in such well-known versions 
as the stories of Beowulf and Grendel, Marduk and Tiamat, Re 
and Apophis, or such less noted versions as the Ngaju Dayak myth 
of the conflict of Hornbill and Watersnake, 7 while the cattle- 
raiding theme is known among all people who keep cattle, ap- 
pearing in such versions as the struggle of Nuer and Dinka, Masai 
and Kikuyu, or David's raids while living among the Philistines. 8 

Both of these themes were well known to the Indo-Europeans. 
Thorr and the MidgarQ serpent, Herakles and the Hydra, 
Karasaspa and the horned serpent are but a few of the I-E monster- 
slaying stories, 9 while the cattle-raiding theme can be seen in such 
tales as Indra and the Panis, Queen Medb's quest for the Bull 
of Cualnge, or Odysseus's men and Helios's cattle. 10 The great 
number and wide distribution of these stories show that they were 
highly popular themes among the I-Es, but in all probability none 
is of great antiquity ascending to the Proto-Indo-European 
period. Rather, they are all simply endless variations on a much- 
beloved theme. They are Greek myths or Celtic myths or Indian 
or Iranian myths, but none of these has a proper claim to Indo- 
European status. 11 

This is not the case, however, with the *Trito myth, which, 
as we shall see, is simultaneously a myth of victory over a ser- 
pentine monster and of the winning of cattle. It is clearly attested 
in India, Iran, Greece, and Rome, as has long been recognized. 

7 For Beowulf, see the translation in G. N. Garmonsway et al., Beowulf and Its 
Analogues (New York: E. P. Dutton & Co., 1969), with its appendix on other 
Norse and Celtic dragon -slaying stories (pp. 333-39); for Marduk and Tiamat, 
Alexander Heidel, trans., The Babylonian Genesis (Chicago: University of Chicago 
Press, 1951); for Re and Apophis, translation by John A. Wilson in James B. 
Pritchard, ed., Ancient Near Eastern Texts, 3d ed. (Princeton N.J.: Princeton 
University Press, 1969), pp. 6-7; for Hornbill and Watersnake, Hans Scharer, 
Ngaju Religion (The Hague: Martinius Nijhoff, 1963), esp. pp. 163-203. 

8 For Nuer and Dinka, see E. E. Evans-Pritchard, Nuer Religion (Oxford: 
Clarendon Press, 1956), pp. 11 ff.; for Masai and Kikuyu, M. Merker, Die 
Masai (Berlin: Dietrich Reimer, 1904), p. 196; on David among the Philistines, 
1 Sam. 27:8-12. 

9 Numerous materials have been collected by Joseph Fontenrose in Python 
(Berkeley: University of California Press, 1959), though I cannot agree with his 
interpretations. 

10 No adequate collection of the great number of I-E myths and legends of 
raiding has yet been attempted, but note Josef Weisweiler, "Vorindogermanische 
Schichten der irischen Heldensage," Zeitschrift fur Celtische Philologie 24 (1954): 
27-28, and A. Venkantasubbiah, "On Indra's Winning of Cows and Waters," 
Zeitschrift der deutschen morgenlandischen Gesellschaft 115 (1965): .120-33, for 
some of the important Irish and Indian references. Again, the interpretations of 
both scholars leave something to be desired. 

11 One of "the most interesting of these recent myths is that of Odysseus and 
Cyclops, which is a myth that combines monster slaying and cattle stealing. On 
the etymology of Cyclops as "cattle -thief," see Paul Thieme, "Etymologische 
Vexierbilder," Zeitschrift fur vergleichende Sprachforschungen 69 (1951): 177-78. 

44 



History of Religions 

Baltic and Slavic parallels have also been cited 12 and a Hittite 
version may also be discerned, though there are numerous diffi- 
culties in this case; 13 an Armenian version is known, though this 
ultimately depends on the Iranian version, 14 and a Germanic 
version is also clear, though this has not heretofore been recognized. 
The relations between the various versions can be schematized, 
as in figure 1. 

For the present, I shall not attempt to deal with all of these 
variants but will take up only those which are most helpful for a 
reconstruction of the P-I-E version. Given the controversy sur- 
rounding the Hittite version, it is best omitted at this stage of the 
study, and as the Baltic and Slavic reflexes exist mainly in the 
form of folktales recorded at a very late date, they present many 
problems of interpretation that are best left to experts in that 
field. 15 Also, the independent Roman version of the myth — the 
story of the Horatii and the Curiatii of Livy 3. 12 — has been al- 
most entirely historicized, and many of the elements which I take 
to be crucial have been lost. While this version provides a mar- 
velous example of the way in which the Roman mentality trans- 
formed Indo-European myths, 16 it does not really help us in a 

12 See V. Ivanov and V. Toporov, "Le Mythe Indo-Europeen du dieu de l'orage 
poursuivant le serpent: Reconstruction du schema," in Echanges et communi- 
cations: Melanges offer ts a Claude Levi-Strauss (The Hague: 1968), pp. 1180-1206. 
Also note the story of the third brother who rescues princesses and treasures 
from a monster in the underworld, which Gubernatis recognized as related to the 
♦Trito myth, in W. R. S. Ralston, Russian Folktales (London: Smith, Elder & 
Co., 1873), pp. 73-84. The variant of Ivan Suchenko (p. 84) is particularly sig- 
nificant, as the monster who is slain appears as three serpents. 

13 This myth, which appears in translation by Albrecht Gotze in Pritchard, 
pp. 125-26, has been the subject of great debate. Paul Kretschmer championed 
the I-E origin of the myth in his articles, "Weiteres zur Urgeschichte der Inder," 
Zeitschrift fur vergleichende Sprachforschungen 55 (1927): 78-79, and "Indra und 
der Hethitische Gott Inaras," Kleinasiatische Forschungen 1 (1927): 297-303. 
Some of his philological arguments, however, may no longer be accepted. Others 
who saw an I-E myth in this text were Giuseppe Furlani, La Religioni degli 
Hittiti (Bologna: Nicola Zanichelli, 1936), p. 42; Friedrich Hrozny, "Hethiter 
und Inder," Zeitschrift fur Assyriologie 38 (1928): 184-85; and Jean Pryzluski, 
"Inara et Indra," Revue Hittite et Asianique 5 (1940): 142-46. Those who attacked 
this conclusion were Ferdinand Sommer, Die Ahhijdva Urkunden (Munich: 
Bayerischen Akademie der Wissenschaften, 1932), pp. 22-24, 383; Albrecht 
Gotze, Kulturgeschichte des alten Orients, II, 1.3 Kleinasien (Munich: C. H. Beck, 
1933), p. 131 n.; and Emmanuel LaRoche, Recueil d^onomastique Hittite (Paris: 
C. Klincksieg, 1951), p. 95. 

14 See Herman Lommel, Der Arische Kriegsgott (Frankfurt: V. Klostermann, 
1939), pp. 51-53; Georges Dumezil, "Vahagn," Revue de Vhistoire des religions 
117 (1938): pp. 152-70. 

15 For literature on the Hittite version, see n. 13 above. With regard to the 
Baltic-Slavic reflexes, the best treatment to date is that of Ivanov and Toporov, 
where many of the same features which will be discussed in this paper are brought 
out: the serpentine nature of the enemy, the importance of cattle as booty, etc. 

16 Georges Dumezil, Horace et les Curiaces (Paris: Gallimard, 1942), pp. 89- 
140, and Destiny of the Warrior, pp. 12-28. 

45 



Indo-European Cattle-raiding Myth 

Proto-Indo-European 



Hittite Indo -Iranian Balto- Slavic Greek Roman! Germanic 




Indian 2 Armenian 



Fig. 1. — Variants of the myth. Indian j = original version, Trita and Visvarupa; 
Indian 2 = remodeled version, Indra and Vrtra; Roman! = independent version, 
Horatius and the Curiatii; Roman 2 = dependent version, Hercules and Cacus. 

reconstruction of the original. Finally, the later Indian develop- 
ments, in which the combat of Indra and Vrtra was modeled on 
that of Trita and Ahi (or Visvarupa), 17 do not really provide us 
with much in the way of new material. Thus we will focus on the 
remaining versions: Indian (Trita and Visvarupa/Ahi), Iranian 
(Thraetaona and Azi Dahaka), Armenian (Vahagn and the dragon), 
Greek (Herakles and Geryon), dependent Roman (Hercules and 
Cacus), and Germanic (iconographic only). The Indian and Iranian 
texts provide a convenient starting point: 

RG VEDA 10.8.8-9 

Aptya, knowing the ancestral weapons and impelled by Indra, did 

battle. 
Having killed the three-headed, seven-bridled one, Trita drove off the 

cattle of Tvastr's own son. 
The mighty lord Indra struck down the conceited one who had sought 

great power. 
Driving forth the cattle of Visvarupa, Tvastr's own son, he ripped off 

those three heads. 18 

17 On the secondary nature of Indra-Vrtra, see the classic treatment of Emile 
Benveniste and Louis Renou, Vrtra et Vrdragna (Paris: Imprimerie Nationale, 
1934), esp. pp. 93-94. 

18 sa pitryany ayudhani vidvan indresita aptyo abhy ayudhyat / 
trisirsanam saptarasmim jaghanvan tvastrasya cin nih sasrje trito gah // 
bhurid indra udinaksantam ojo 'vabhinat satpatir manyamanam / 
tvastrasya cid visvarupasya gonam acakranas trini sirsa para vark //. 

46 



History of Religions 

yast 15.23-24 

Thraetaona, the son of the house of Athwya, of the heroic house, sacri- 
ficed to him [i.e., Vayu] in four-cornered Varana, on a golden chair, on a 
golden pillow, on a golden rug. He strewed the sacred twigs from cupped 
hands streaming [libations]. 

He asked: "Grant success to me, O Vayu, whose deeds are the highest, 
that I might be victorious over Azi Dahaka, the three-mawed, three- 
skulled, six-robed, strong daevic lie possessed of a thousand powers, the 
evil betrayer of mankind, who is the strongest lie created by Anra Mainyu 
[the Evil Spirit] against the material world and to destroy the world of 
Asa [Right]. And may I carry off his two women [vantd], Savanhavac and 
Aranavac, who raised themselves up with the most beautiful bodies for 
the world, who are the most excellent." 19 

Both of these texts describe the slaying of a three-headed 
monster. Moreover, the name of the hero in both cases is the same, 
for the Indian Trita Aptya corresponds perfectly to the Avestan 
Thraetaona Athwya. Their first names, while not identical, are 
closely related, Thraetaona being a patronymic from Avestan 
Thrita, which is a perfect match for Sanskrit Trita, both being 
derived from P-I-E *Tri-to-. This *Tri-to- has further reflexes: 
Greek Tplros, Old Norse pridi, Albanian trete, Latin tertius, Welsh 
trydydd, Old High German dritto, and so forth, and means literally 
"third." 20 The second name, *Atpya, seems simply to be the 
name of a family of Indo-Iranian heroes 21 and, contrary to the 

19 torn yazata viso puQro aQwyanois viso siiraya Qraetaono upa varonom 
caOru.gaosom zaranaene paiti gatvo zaranaene paiti fraspaiti zaranaene paiti 
upastarone frastarotat paiti barosman poronaeibyo paiti yzarayatbyo. aom 
jaioyat avat ayaptom dazdi me vayo yo uparo.kairyo yat bavani aiwi.vanya azim 
dahakom Orizafanom 0rikamoro8om xsvasasim hazanra.yaoxstim asaojanhom 
daevimdrujom ayom gaeOavyo drvantom yamasaojastamamdrupm fraca korontat 
arjro mainyus avi yam astvaitim gaeGam mahrkai asahe gaeGanam. uta he vanta 
azani savaqhavaca oranavaca yoi horn kohrpa sraesta zazaitoe gaeGyaica yoi 
abdotome. Translation of the last phrase following Karl Hoffmann, "Junga- 
westisch zazaite," Munchener Studien zur Sprachwissenschaft 4 (1954): 45-52. 

20 Julius Pokorny, Indogermanisches Etymologisches Worterbuch (Bern: Francke 
Verlag, 1969), p. 1091; Manfred Mayrhofer, Kurzgefasstes etymologisches Worter- 
buch des Altindischen (Heidelberg: Carl Winter, 1956), 1 : 534-35; Hermann Giintert, 
Der arische W eltlconig und Heiland (Halle: Max Niemeyer, 1923), p. 31; Dumezil, 
Destiny of the Warrior, p. 14; Walther Wiist, "Trita und Verwandtes," Worter 
und Sachen 3 (1940): 225-27. Kasten Ronnow, Trita Aptya, eine vedische Gottheit 
(Uppsala: Appelberg, 1927) attempted to show a relation to Greek rpircov, both 
being water deities, but the long -i- of rplrcov makes this impossible (see Hjalmar 
Frisk, Griechisches Etymologisches Worterbuch [Heidelberg: Carl Winter, 1960], 
2: 933-34). The line of argument suggested by Jacob Wackernagel, "Akzentstudien 
I," Nachrichten der Gesellschaft der Wissenschaften, Gottingen, Philosophisch- 
Historisch Klasse (1909), pp. 60-61, n., accepted by Herman Lommel, "Naotara 
und Spitama," Indogermanische Forschungen 53 (1935): 183, and Jacques 
Duchesne-Guillemin, "Aw. Graetaona-," Indogermanische Forschungen 54 (1936): 
205, is labored and unneccessary. 

21 Alfred Hillebrandt, Vedische Mythologie (Breslau: M. & H. Marcus, 1902), 
3: 341-43. This remains the most probable explanation, though not the most 
widely accepted. In any event, there are no further Indo-European corres- 
pondences, and the problem involves only interpretation of material specific to the 
Indo-Iranian context. 

47 



Indo-European Cattle-raiding Myth 

generally accepted opinion, has nothing whatever to do with water, 
as Bartholomae and Wackernagel have convincingly demon- 
strated. 22 

The only mutation which needs to be accounted, for, then, 
is the shift from Indo-Iranian *Trita- to Avestan Thraetaona, 
"Son of Thrita." Why should we have this slip in generations? 
The answer seems to be found in the Iranian ideology of the 
making of a hero. To judge from certain Avestan texts, most 
particularly Yasna 9, heroism was something of a two -generation 
affair for the Iranians. It is part of a cycle bound up with the 
reformed haoma sacrifice, whereby if a father wishes heroic sons 
he prepares the intoxicating haoma according to ritual, offers 
some to the gods, and drinks some himself. Once within his body, 
it descends to the genitals where it is distilled into semen, and 
this, when transmitted to a son, develops into the x v ardnah, the 
radiant nimbus enveloping heroes and kings. 23 This complex 
theory is a specifically Iranian piece of speculation, and in India 
one who prepares and drinks the soma ( = haoma, both being 
derived from I-I *sauma- "pressed drink") is himself gifted with 
heroic properties without having to wait a generation. 24 Now 
Trita and Thrita are both known as preparers of *sauma, as in 
RV 8.12.16, 1.187.1, or Yasna 9.10, and it thus seems likely that 
in the original Indo-Iranian version *Trita prepared and drank 
the *sauma himself, thus taking on heroic powers. The develop- 
ment whereby Thrita — > Thraetaona is a later development, and 
the Avestan Thrita and Thraetaona can both be understood as 
reflexes of an I-I *Trita. 

It should be noted that this *Trita was a mythic personage 
and a great hero, but certainly was not a god. There is no evidence 
on the Iranian side for divine status, 25 and the majority of the 

22 Christian Bartholomae, Arische Forschungen (Halle: Max Niemeyer, 1882), 
1:8 n. & f., and "Arica I," Indogermanische Forschungen 1 (1892): 180-82; 
Wackernagel, p. 61 n., and a note contributed to Hanns Oertel, The Syntax of 
Cases in the Narrative and Descriptive Prose of the Brahmanas (Heidelberg: Carl 
Winter, 1926), p. 328. 

23 Thus, in Yasna 9, each of the first four men to prepare Haoma are themselves 
not heroes but are blessed with heroic sons: Vivanhant with Yima, A0wya with 
Thraetaona, Thrita with Karasaspa, and Pourusaspa with Zarathustra. All of 
the fathers prepare Haoma and all of the sons bear the x v ardnah. On the nature 
of x v ardnah and its relation to mystic physiology, see J. Duchesne-Guillemin, 
"Le X v aranah," Annali Istituto di Napoli, sezione linguistica 5 (1963): 25-26, 
and, in a broader context, Mircea Eliade, "Spirit, Light and Seed," History of 
Religions 11 (1971): 1-30. 

24 As, for instance, in the famous hymn of the man drunk with soma, RV 
10.119. 

25 Roth, pp. 220-21, 225. His reporting of the evidence is perfect, but he came 
to the mistaken conclusion that Trita was originally a god who sank to human 
status in Iran while retaining his original form in India. 

48 



History of Religions 

forty-odd occurrences in the Rg Veda show him to be human 26 — 
in contact with the gods and aided by them, but no god himself — 
thus, for example, RV 5.86.1, 10.48.2, 2.11.19, 9.34.4, 2.34.14, 
10.64.3, 8.52.1, and so forth. Those verses where he does occasion- 
ally appear as a god, such as RV 5.41.4 and 8.41.6, are either the 
result of independent Indian elevation of the hero or, perhaps 
even a second Trita. 27 Originally, though, like *Manu and *Yemo 
with whom he is often grouped, 28 *Trito was one of the first men 
in myth whose actions helped to shape the world and continue to 
serve as a prototype for men of the present day. 

Given the Indo-Iranian correspondence, plus the other I-E 
correspondences which we will take up below, numerous scholars 
have been inclined to treat the Indian Trita as the original figure 
of the serpent- or dragon-slaying myth, seeing Indra — who often 
plays the role of dragon slayer — as a late intruder into the cycle. 29 
In their opinion, in the earliest versions of the myth it is Trita 
and 'Trita alone who slays the monster, Indra taking this role 
only in later versions. Those versions in which the two appear 
together as in the text we have cited are seen as cases of incom- 
plete dominance, composed before Indra had completely over- 
shadowed his forerunner. This theory, however, does not seem 
to be supported by the evidence. Rather than a gradual process 
of eclipse, there is a consistent relation of assistance and depen- 
dence between Indra and Trita. In the text under consideration, 
both figures are credited with having killed the enemy, and Trita 

26 Hillebrandt, p. 344; Herman Oldenberg, Die Religion des Veda (Berlin: 
Wilhelm Hertz, 1894), pp. 143-44. Others have had varying opinions, and 
attempts have frequently been made to identify Trita with a god. Thus, Abel 
Bergaigne, La Religion Vedique (Paris: Honore Champion, 1963), 1:328-9, 
identified him with Apam Napat; A. A. Macdonell, "Mythological Studies in the 
Rigveda: The god Trita," Journal of the Royal Asiatic Society ( 1893), pp. 419-96, and 
Murray Fowler, "Trita Soter," Journal of the American Oriental Society 67 (1947): 
59-60, saw him as Agni; Edward Delavan Perry, "Indra in the Rig- Veda," 
Journal of the American Oriental Society 11 (1885): 142-48, and Roth saw him as a 
storm god; Ronnow saw him as a water god; L. D. Barnett, "The Genius: A Study 
in Indo-European Psychology," Journal of the Royal Asiatic Society (1929): 
740—41, saw him as the father of the Maruts. The very diversity of opinion might 
indicate the mistaken nature of the quest. 

27 Friedrich Spiegel, Arische Periode und ihre Zustdnde (Leipzig: Wilhelm 
Friedrich, 1887), pp. 268-69, suggested this, and the appearance of a Trita 
Vaibhuvasa in RV 10.46.3, in contrast our Trita Aptya, lends credence to the 
suggestion. 

28 See RV 10.64.3, 1.163.2-3, 8.52.1. In Iran, see Yasna 9.3-13 and Yast 
19.36. 

29 Thus: Jarl Charpentier, Kleine Beitrdge zur indoiranischen Mythologie 
(Uppsala: Akademische Buchdruckerei, 1911), p. 56; A. B. Keith, Religion and 
Philosophy of the Vedas and Upanishads (Cambridge Mass., Harvard University 
Press, 1925), pp. 127-28; Oldenberg, p. 143; Perry, p. 144; Leopold von Schroeder, 
Mysterium und Mimus in Rigveda (Leipzig: H. Haessel, 1908), p. 132; R. N. 
Dandekar, "Vrtraha Indra," Annals of the Bhandarkar Oriental Institute 31 
(1950): 10-11. ' 

49 



Indo-European Cattle-raiding Myth 

is explicitly said to be "impelled by Indra" (indresita), being simi- 
larly aided by him in RV 10.48.2, 5.86.1, and 2.11.19. Trita, in 
turn, gives soma to Indra (RV 9.34.4, 9.86.20) and is said to drink 
the intoxicating brew alongside the god in RV 8.12.16. All in all, 
their relation might be explained as an exchange of strength, 
whereby the warrior hero offers sacrifice of *sauma to the warrior 
god. This drink strengthens the latter and allows him to bring 
strength and assistance to the hero in return. Given this, we are 
led to see both Trita and Indra as originally present in the Indian 
version of the myth, Trita appearing as the hero who slew the 
monster and Indra as the god who aided in the exploit. 

The Iranian evidence is somewhat inconclusive on this point. 
In the text cited above, Thraetaona calls on the god Vayu for 
help, and in parallel passages elsewhere he similarly calls on Ardvi 
Sura Anahita (Yt. 5.33-34), Drvaspa (Yt. 9.13-14), and Asi 
Varjuhl (Yt. 17.33-34). Clearly the hero was aided by a deity in 
Iran also, the only question being which one was original. Given 
the changes which the reform of Zarathustra effected in the status 
of *Vrtraghna *Indra, 30 I am led to hypothesize that this old 
Indo-Iranian warrior god originally played this role, and that only 
after his demotion at the prophet's hands did other deities rush 
in to fill the vacuum. 31 But we need not content ourselves with 
hypothesis, for there is another text that allows us certainty on 
this point. 

MOSES OF CHORENE "HISTORY OF ARMENIA" 1. 31 

His [Tigran's] sons were Bab, Tiran, Vahagn: of this latter, the fables 
tell: "Heaven and earth were in travail, the purple sea was in travail; a red 
reed had its birth in the seas, from the stems of the reed came forth smoke, 
from the stems of the reed came forth a flame, and from the flame sprang 
a young man; this youth had fiery hair, also a beard of flame, and his eyes 
were suns." All sing of this one, I have heard it with my own ears; they thus 
recount in song along with cymbals, his battle with the dragon and his 
victory, and they sing of him in every way as of the heroic deeds of Her- 
cules. 32 

The first point which we must note is, as has long been known, 
the name Vahagn is a loan word into Armenian and is derived 

30 Following Benveniste-Renou, I take *Vrtraghna to be the original name of 
the Indo-Iranian warrior god. But, as the name Indra appears in Avestan 
(Vendidad 10.9 and 19.43) as well as Sanskrit, this must also be accepted as an 
I -I epithet of the god. 

31 As argued by Geo Widengren, Die Religionen Irans (Stuttgart: W. Kohl- 
hammer, 1965), p. 18; R. C. Zaehner, The Dawn and Twilight of Zoroastrianism 
(New York: G. P. Putnam's Sons, 1961), p. 89; Dumezil, Destiny of the Warrior, 
pp. 116-17. 

32 Translation following the German of M. Lauer, Der Moses von Chorene 
Geschichte Gross Armeniens (Regensberg: Georg Mainz, 1869), p. 52. 

50 



History of Religions 

from Avestan Vdrddrayna ( = Skt. Vrtrahan). 33 The story of his 
birth from a flaming reed has been connected with an Indian 
itihdsa tradition telling of the reenergizing of Indra. 34 Thus we are 
virtually certain that Vahagn is a dependent variant of the Indo- 
Iranian warrior god, who was known as both *Vrtraghna, "smasher 
of resistance," 35 and *Indra, "the manly, the strong." 36 Ac- 
cordingly, his role as a dragon slayer in this text has frequently 
aroused the interest of scholars. 37 Two other details should be 
noted as well. First, Vahagn is set in a distinctly human genealogy, 
descended from a king and having two human brothers. He is 
compared with a hero, Hercules, and not to a god. 38 Thus, despite 
his bearing a name derived from that of the warrior god, Vahagn 
is a mortal. Second, Vahagn is the third child born to Tigran. 
The other two play no roles of any importance and seem to have 
been added simply to preserve Vahagn's position as third in line. 39 
So he seems to represent not only the god *Vrtraghna, but also 
the hero *Trito, "Third," the roles of god and hero who originally 
cooperated in the dragon slaying having fused into one figure, a 
hero who bears the name of the god. 

In the Armenian version, however, the myth is given in a most 
abbreviated form. As a result, *Trito's enemy has lost much of the 
specificity he had in the Indian and Iranian versions. There, in 
the first place, he is described as a serpent, Avestan azi, Sanskrit 
ahi (see RV 10.48.2), both of which are derived from P-I-E 
*ng w hi-, "serpent," of which Latin anguis, Lithuanian angis, 
Armenian auj, Greek oyis, and Middle Irish esc-ung, "eel" (lit. 

33 Benveniste-Renou, p. 75. 

34 Dumezil, "Vahagn," pp. 152-70. 

35 Benveniste-Renou, pp. 19-22. 

36 On the etymology of the name Indra, now see H. W. Bailey, "The Second 
Stratum of the Indo-Iranian Gods," in Mithraic Studies, ed. John R. Hinnells, 
2 vols. (Manchester: University of Manchester Press, 1975), 1: 9-11 and n., which 
serves to vindicate the view of H. Jacobi, "Uber Indra," Zeitschrift fur verg- 
leichende Sprachforschung 31 (1892): 316-19. 

37 Thus Lommel (pp. 51-53), Zaehner (p. 103), and others have insisted that 
since Vahagn was said to be a dragon slayer, the Iranian VaraOrayna must have 
played this role. In anticipation of their line of argument, Benveniste (Vrtra et 
Vrdragna, pp. 75 ff.) was led to make some very ill-founded remarks, attempting 
to derive Vahagn from Herakles in his deeds and VaraOrayna in his name. 

38 There is still further identification of Herakles and the I -I *Trita in other 
sources. Thus, in relating Scythian legends, Herodotus 4. 8-10 places Herakles 
in the role of Feridiin ( < Thraetaona) in the story of the initiation of his three 
sons. For the Middle Persian version, see Reuben Levy, trans., The Epic of the 
Kings: Shah-nama (Chicago: University of Chicago Press, 1967), pp. 26-27. 

39 For similar transformations, note that the three-headed Geryon is turned 
into three brothers in Diodorus Siculus 4. 17. 2, and that the hero *Trito becomes 
the last of the three Horatii brothers in Livy 1. 25-26. On this latter, see Dumezil, 
Horace et les Guriaces, pp. 89-140, or Destiny of the Warrior, pp. 12-28. 

51 



Indo-European Cattle-raiding Myth 

"serpent of the waters") are also reflexes. 40 Second, this serpent 
is said to be three headed, Avestan 6ri-kamdrdddm, Sanskrit 
tri-sirsdnam. Finally, this three-headed serpent is further marked 
as a *ddsa-, an aboriginal inhabitant who is inimical to the Indo- 
European invaders. The Iranian Azi Dahaka carries this in his 
name, the -ka apparently being a prejorative suffix, 41 while 
Visvariipa is described as a ddsa in RV 10.99.6 and 1.158.4. 

One element is completely lacking in the Armenian version, 
and in my opinion it is a crucial one: the booty won in the en- 
counter. Moreover, our Indian and Iranian sources leave some 
ambiguity on this point, for while the Indian story of Trita's 
victory states that cattle were the plunder, 42 the Iranian version 
tells how two women previously taken from Yima by Azi Dahaka 
were won back by Thraetaona. 43 Given this set of facts, some 
scholars have been led to see both "cattle" and "women" as 
symbolic forms referring back to natural phenomena, specifically 
the storm or the seasonal freeing of the waters. 44 The myth is 
taken as if allegory, *Vrtraghna and *Trita being identified with 
the storm, *Aghi (the I-I form < I-E *ng w hi-) with the clouds, 
and the cows or women with the rain. While the myth may have 
taken on this allegorical coloring in some variants under the im- 
pact of later Indian speculative thought, 45 it is doubtful that this 
is the original meaning. Rather, the alternation between cows and 
women can be explained in quite another fashion. 

In order to appreciate this, it is instructive to look at the 
specific term used to describe the women won by Thraetaona, 

40 Pokorny, pp. 43-44, with modification of the initial vocalism. 

41 On ddsa, see Mayrhofer, 1:28-29. 38 f.; Emile Benveniste, Vocabulaire des 
institutions Indo-Europeens (Paris: Minuit, 1969), 1:369. The -ka suffix in Dahaka 
seems analagous to that in the Sanskrit demon of the Puranas, Dhenu-ka, "Little 
(evil) cow." 

42 RV 10.48.2, 10.8.8. While other booty does appear in the stories of Indra's 
various combats, Trita never wins anything other than cattle. Even with regard 
to these later stories of Indra's exploits, the winning of cattle is a consistent and 
pervasive feature, as Venkantasubbiah has shown. 

43 See the analysis of James Darmesteter, Le Zend Avesta (Paris: Maisonneuve, 
1960), l:xlvi-xlviii. 

44 Macdonnell, p. 467 et passim; Perry, p. 144; Roth, pp. * 223-24; von 
Schroeder, Mysterium und Mimus, pp. 132 ff., and Herakles und Indra, pp. 
57-67; Breed, pp. 1-162; Widengren, pp. 41-42. 

45 Yaska, Nirukta 2.6, states that in his time there were already two schools 
of thought as to the identity of Vrtra in the Indra-Vrtra myth: some held him 
to be a demon, others, a cloud. Sayana, following the latter, consistently inter- 
preted the "cows" won in the battle as rain, and it is largely due to the early 
Western overreliance on Sayana for interpretation of the Veda that this notion 
of the storm allegory came to such prominence in the work of the "comparative 
mythology" school. As these speculations are not found outside of India, though, 
we may be certain that they are the product of the Indian speculative genius, 
and do not ascend to the I-I or I-E period. 

52 



History of Religions 

Avestan vantd. Bartholomae, following Darmesteter's line of 
investigation, glossed this word as "die Geliebte, Frau." 46 But 
when one analyzes the word, it is clear that it is nothing more than 
a feminine form of a past passive participle of the verb Vvan- 
"to wish for, desire," as Bartholomae himself noted. 47 Thus, in 
reality it means no more than "the female who is desired." Such 
a term could surely apply to bovines as well as to humans under 
certain circumstances. 

A similar term is Indo-Iranian *dhainu- (Skt. dhenu- = Av. 
daenu-), one of the most frequent terms for "cow." Yet, as Ben- 
veniste has shown, the word means nothing more than "one who 
lactates, gives milk," being derived from the verb "to give milk, 
nourish" (Skt. Vdhai-)* 8 As such, it may be used for the female 
of any species, Homo sapiens included, 49 and in a very important 
verse from the Rg Veda (5.30.9) the parallel term dhend, usually 
rendered "cows," is used to describe two women who have been 
captured by Dasa enemies. 50 

The Iranian version of the myth has been well worked over, 
as is clear in our text. In general, a historicizing tendency is 
obvious, as the struggle between hero and serpent has become a 
dynastic dispute, the monster having become a human — albeit 
three-headed — usurper. 51 It seems highly probable, then, that 
the ambiguity of the Avestan words vantd and daenu allowed for 
a rationalization whereby a myth of cattle raiding became the 
myth of recovering abducted queens. 

Such a formulation finds support in the Greco-Roman versions 
of the myth. These variants and their relation to the Indian 
Indra-Trita lore were among the first texts to be dealt with by 
the nineteenth-century Indo-Europeanists of the "Comparative 
Mythology" school. Often the texts were abused, and nature 
allegory was imposed on every detail. 52 As a result of this and 

46 Christian Bartholomae, Altiranisches Worterbuch (Berlin: de Gruyter, 1961), 
col. 1355. 

47 Ibid., note. 

48 Benveniste, Vocabulaire, pp. 22-23. 

49 Ibid. 

50 Widengren, pp. 46-47, has called attention to this text, though for a very 
different purpose. 

51 Thraetaona's constant motive is to throw out A£i, who is variously seen 
as a Dacian or Turanian, and to restore the proper Iranian rule that ended with 
Yima's fall. He thus becomes the national Iranian hero par excellence, while 
Zohak ( < Aii Dahaka) is remodeled to fit changing historical situations, becoming 
an Arab in the Shah-nameh (see Levy, p. 12). 

52 Von Schroeder, Herakles und Indra, pp. 57-67, is perhaps the best example 
of such Procrustean interpretation. 

53 



Indo-European Cattle-raiding Myth 

Dumezil's striking demonstration that quite an unexpected 
Roman source contained an independent and very different reflex 
of the myth, 53 these versions which had been recognized earlier 
fell into disuse, though the validity of their comparison with Indian 
and Iranian traditions has never been soundly challenged. They 
are extremely important for our case, since they have clearly 
and unambiguously preserved the fact that in the mythic en- 
counter between a hero and a three-headed monster cattle were 
the prize that was sought and won by the hero. 

beesiod "theogony" 11. 287-94 

And Khrysaor begat the three-headed Geryon, 

Being married to Kallirhoe, the daughter of famed Okeanos. 

Now the force of Herakles killed him 

Beside the shuffling cattle in sea-girt Erytheia 

On that very day when he drove the broad-faced cattle 

To holy Tirynth, having passed over the ford of Okeanos, 

And having killed Orthos and the cow -herd Eurytion 

In the gloomy herdsman's house beyond famed Okeanos. 54 

propertius "elegies" 4. 9. 1-20 

In that season when Amphitryon's son bore off 

The young oxen from your stalls, O Erythea, 

He came to the Palatine hill which is unconquered by man, 

And himself being weary, he set down his weary cattle 

Where Velabrum overflows its stream and where 

The seaman sails through urban waters. 

But they did not remain safe with Cacus, an unfaithful 

Host: that one defiled Jove with theft. 

Cacus was a native-dweller [incola], a robber from a dreaded 

cave, 
Who uttered sounds through three separate mouths [tria 

partitos . . . or a], 
This one, in order that there would not be any sure clues 

giving signs of the robbery, 
Dragged the cattle backward by the tail into his cave 
But not without witness by the god: the young oxen betrayed 

the thief with their sounds, 
And wrath pulled down the rough doors of the thief. 
Cacus lay dead, struck three times by the Maenalian bough, 
And Alcides spoke thus: "Go, you cattle, 

53 Dumezil, Horace et les Curiaces. 

Xpvadcop 8' €T€K€V rptK€<paXov rrjpvovrja 
fieixOcls KCtAAtporj Kovprj kXvtov *QK€avoio' 
rov p.kv dp* €$€vdpi$e f$lr) 'HpaKXyeirj 
j3ovai Trap eiXnrohtaoi Trepippvra) elv 'Epvdcir), 

iJ/ZCCTl TO> OT€ 7T€p f$OVS TJXaaCV €VpVfJL€T(l)7TOVS 

Tipvvd* els Uprjv Siafias Tropov 'QKeavoio, 

"Op0OV T€ KTClVaS KGU fioVKoXoV EvpVTlCUVCt 

aradfia) ev rjipoevri Tript]v kXvtov 'QKtavoZo. 

54 



History of Religions 

Go, you cattle of Hercules, the last labor of our club! 
Twice sought by me, and twice my booty, you cattle — 
Sanctify the fields and Cattle -place with your long lowing. 
The Forum of noble Rome will be your pasture." 55 

Now these are not fully independent versions of the myth. 
The Greek version of Hesiod is apparently independent, but 
Propertius and the other Romans who told this story all seem 
to rely on the Greek sources in large measure. 56 The Greek text, 
while very sparse, does give us some valuable information, 
especially in this translation. In order to appreciate this it is 
crucial to note that, contrary to the general opinion since an- 
tiquity, 57 Hesiod did not take Herakles to be the slayer of Orthos 
and Eurytion; rather, this role was allotted to Geryon. The 
matter hinges on the question of who is the subject of line 291. 
Grammatically, the verb rjXavev has only an implied subject, 
which may refer back to Geryon (mentioned in the accusative 
in line 287, and referred to by the pronoun tov in line 289) or 
Herakles (mentioned in the dative in line 289). However, the fact 
that Herakles occurs only in an oblique case (jS/^ being the sub- 
ject of line 289) makes it less likely that he is meant as the sub- 
ject here. But what to my mind is most persuasive is the presence 
of the emphatic particle rrep in line 291, giving the sense of "on 
that very same day" to the phrase rj^arc to) ore -rrep. If I am right, 
the motivation behind this phrase is the attempt to show that two 

55 Amphitryoniades qua tempestate iuvencos 
egerat a stabulis, o Erythrea, tuis, 

venit ad invictos pecorosa Palatia montes, 

et statuit fessos fessus et ipse boves, 

qua Velabra suo stagnabant flumine quaque 

nauta per urbanas velificabat aquas. 

sed non infido manserunt hospite Caco 

incolumes: furto polluit ille Iovem. 

incola Cacus erat, metuendo raptor ab antro, 

per tria partitos qui dabat ora sonos. 

hie, ne certa forent manifestae signa rapinae, 

aversos cauda traxit in antra boves, 

nee sine teste deo: furem sonuere iuvenci, 

fur is et implacidas diruit ira fores. 

Maenalio iacuit pulsus tria tempora ramo 

Cacus, et Alcides sic ait: "Ite boves, 

Herculis ite boves, nostrae labor ultime clavae, 

bis mihi quaesitae, bis mea praeda, boves, 

arvaque mugitu sancite Bovaria longo: 

nobile erit Romae pascua vestra Forum." 

56 Jean Bayet, Les origines de VHercule Romain (Paris: E. de Boccard, 1926), 
p. 233; Michael Grant, Roman Myths (New York: Charles Scribner's Sons, 1971), 
p. 41. 

57 The story is presented this way in Apollodorus Mythographus 2. 5. 10, and 
most later writers seem to have followed his version without questioning whether 
he had properly understood Hesiod or not. 

55 



Indo-European Cattle-raiding Myth 

separate actions took place on one and the same day; that is, that 
on the very day Geryon killed Orthos and Eurytion, making off 
with their cattle, he himself was killed by Herakles, who appro- 
priated the animals in turn. Grammatically there is no reason to 
reject this reading. Moreover, it is more in keeping with the I-E 
background of the myth, for, if it is accepted, it will be seen that 
(a) Herakles is the third opponent to face Geryon, and (b) Geryon 
is the original cattle thief, having stolen the animals in an earlier 
encounter, just as Azi Dahaka stole the two women from Yima 
before Thraetaona recovered them. We see more of these themes 
of thirdness and the first theft in the Roman accounts, and there 
it is even possible to perceive that the original name of the hero 
in this exploit was not Herakles, but most probably rplrog, 
"Third." 58 

Of course Herakles is an extremely complicated figure, and I 
am by no means claiming to derive him in his entirety from a 
P-I-E hero named *Trito. What I do claim is that the original 
hero of the Geryon slaying bore the name "Third" and is derived 
from this P-I-E figure. Herakles himself is a separate entity, a 
hero who for one reason or another became the most popular 
of the Greek world, in the process subsuming the deeds of many 
other heroes, deeds which were organized into a very late and 
artificial cycle of Twelve Deeds. 59 

The Roman Hercules-Cacus story, modeled after Herakles- 
Geryon, has managed to preserve features that point to the origi- 
nal hero's name, much as Hesiod's account preserves the fact that 
Herakles was Geryon's third opponent. The sources which lead 
us to this conclusion are the following: 

a) Propertius Elegies 4. 9. 15 (cited above): Cacus is specified 
as being struck with the Maenalian club three times (tria tempora). 

58 There are some scattered pieces of evidence from the Greek world which 
might point to this same conclusion, though none are particularly convincing in 
and of themselves. Thus, Herakles was commonly called "Herakles of the three 
nights," TpUonepos, due to the story in which Zeus lay with Alcmene for three 
nights in order to conceive him (O. Gruppe, "Herakles," in Pauly Wissowa 
Realenzyklopadie, suppl. 3, pp. 1004, 1016); he is occasionally depicted as three- 
footed on vase paintings (Ernst Curtius, Herakles der Satyr und Dreifussrauber, ein 
griechisches vasenbild [Berlin: W. Hertz, 1852]); in a lost comedy, three Herakleses 
seem to have engaged a triple Geryon in an eating contest (Stephanus Oswiecimski, 
"De tribus Herculibus mimicis, Eos 44 [1950]: 119-22); and in Lucian's Dialogues 
of the Dead, 16.5, there is talk of both a triple {rpi7rXovv) and a third {rpirovl) 
Herakles. It is worth noting that all of these pieces of evidence come from comic 
sources, and it may be that the tradition of Herakles as "Third" was better pre- 
served in this type of literature due to the opportunity for punning and farcical 
representation which it provided. 

59 Bernard Schweitzer, Herakles (Tubingen: J. C. B. Mohr, 1922), p. 146. For an 
intriguing view of the crystallization of the figure of Herakles, see Paul Kret- 
schmer, "Mythische Namen: Herakles," Glotta 8 (1916): 121-29. 

56 



History of Religions 

b) Ovid Fasti 1. 575: Hercules' club, with which he kills Cacus, 
is said to be three-noded (trinodis). 

c) Vergil Aeneid 8. 230 ff.: Having located Cacus's cave, 
Hercules runs around the mountain three times (ter . . . lustrat), 
batters the door three times (ter . . . temptat), and sits down to 
rest three times (ter . . . resedit) before finally breaking in. 

In each of these versions, the number three comes into play, and 
in each it is in a different fashion. Ovid, Vergil, and Propertius 
are not quoting from each other or from some common source, yet 
each seems intent on introducing this numerical detail into the 
story. Further, in each case the detail can be taken as expressing 
the fact that it was the third of some sort of series (blow, node, 
circumambulation, or rush at the door) which caused the monster's 
death. Each author in his own way has preserved the fact that it 
was the "Third" which killed the beast. 

Numerous other features from the Greco-Roman texts show 
resemblance to the versions we have already treated. The hero's 
enemy is a tricephal (rpuKecpaXos; tria partitos . . . ora), and his 
serpentine nature can be discerned in Geryon's genealogy, his 
paternal grandmother being Medusa, the serpent-haired Gorgon. 60 
Cacus, like Azi Dahaka and Visvarupa, is identified as a non- 
Indo-European aborigine (incola), hostile to Greek and Roman 
alike. 61 Moreover, we are told that Cacus first stole the cattle 
which rightfully belonged to Hercules: the latter's acts were only 
justified revenge, a situation corresponding to the facts of the 
Iranian and the Greek version, where the non-Indo-European is 
the original aggressor. 

A split is evident between the Indo-Iranian versions of the 
myth and the Greco-Roman ones with regard to the role of a 
warrior god and libation ritual. For the Indo-Iranians, the help of 
*Vrtraghna *Indra was invaluable in securing victory for *Trito, 
and the hero was careful to pour a *Sauma libation before the battle, 
strengthening the god that he might strengthen him. Yet, in the 
Greco-Roman versions, the hero acts alone and needs no help 
from a divine figure. 62 Moreover, this seems to be a common 

60 Hesiod Theogony 1. 280-81. 

61 Bayet, pp. 208-14. Also note the argument of H. J. Rose, "Chthonian 
Cattle," Numen 1 (1954): 213-27, who saw in the H er aides -Gery on myth and in 
related tales, a myth told by the non-Indo -Europeans to account for the I-E 
invaders possession of cattle. 

62 It is possible that the god and the hero have merged in one figure, the 
demigod Herakles, as they have in the Armenian Vahagn. Yet the name indicates 
that Herakles was originally a mortal, as no other figure in the entire Greek 
pantheon has a name formed from that of another deity, as is ^Hpa-KXcFys 
"Glory of Hera" (H. J. Rose, "Herakles," in Oxford Classical Dictionary, p. 498). 

57 



Indo-European Cattle-raiding Myth 

European state of affairs, for the Germanic version likewise 
preserves no god. 

This version, which to the best of my knowledge has never been 
recognized heretofore, is found in a relief on the golden horn of 
Gallehus, dated from the fifth century a.d. In figure 2 taken from 
this relief, a three-headed man carrying an axe or hammer in his 
right hand leads a horned animal (a goat?) with his left. At his 
side, three serpents lie dead (fig. 2). As always, interpretation of an 
iconographic representation presents many difficulties. The tri- 
cephal with the axe may be the hero *Trito (I lean toward this 
interpretation) or he may be a doublet of the three serpents. One 
cannot be sure. But what is certain beyond any doubt is that this 
is an independent German reflex of our myth, containing the 
themes of triplicity, serpent enemies, and the taking of livestock 
by force. 

We are thus led to reconstruct a myth in which an Indo- 
European hero whose name was *Trito, "Third," suffered at the 
hands of a monstrous figure, a three-headed serpent who was 
explicitly identified with the aboriginals of the area in which the 
myth was told. In the first encounter, this serpent stole some 
cattle belonging to the hero or to someone close to him, but in a 
second meeting (when — according to the Indo-Iranian version — 
the hero was aided by a warrior god and fortified by an intoxi- 
cating drink) he defeated the monster and recovered the cattle. 
The elements which have contributed to this reconstruction are 
laid out in tables 1 and 2. 

We are left with the question of what such a myth can mean. 
Scholars have often interpreted these texts as allegories of the 
storm 63 or of victory over the powers of death. 64 One might also 
be tempted to offer a reading of the myth that sees it as a conflict 
of cosmic realms, the hero coming from above and the serpent 
from below. 65 Psychological explanations have also been at- 
tempted. 66 Yet, in my opinion, such interpretations fail to take 

63 Thus Breal, Schroeder, Macdonnell, etc. 

64 Fontenrose; Schweitzer, pp. 87, 121. 

65 The main problem with such a view is that there are no qualities which 
point toward a celestial identification for the hero. The conflict of eagle and ser- 
pent, well known among the Indo -Europeans, does have such a significance 
(see David M. Knipe, "The Heroic Theft: Myths from Rgveda IV and the Ancient 
Near East," History of Religions 6 [1967]: 328-60). 

66 For a Freudian view, see Geza Roheim, "The Dragon and the Hero," 
American Imago 1 (1940): 40-69; for Jungian, Erich Neumann, The Origins and 
History of Consciousness (Princeton, N.J.: Princeton University Press, 1954), 
pp. 152 ff., or Jung himself, Symbols of Transformation (Princeton, N.J.: Prince- 
ton University Press, 1956), pp. 365-75. 

58 




59 



g 


gq 


TJ 




GQ 


T* 


o 




<D 




o 


<D 


W 


13 


-P 


GQ 


-P 


-P 




o 


-P 


3 
§ 




-P 




w 


o 


O 


H A 


o 



w 

ffl 
H 



H 
o 

fc 
H 
Q 
fc 
O 

CO 

W 

SI 

o 
o 





s 


bo 


CQ 
-P 

© 

© 


3 

a 




© 


o 


H 


H 


m 



TJ T* 



O O 



S 
o 



a 

o 



o 

O 



I 






c8 



H m > H 



1 



W ft 





© 


5 


U 


ti 


X 


w 


H 



J5Q g 

3 12 

S bO 



^2 « 






GQ O 

02& S O 



c8 
O 



c8 



s 




TS 


T* 


T* 


T* 




T* 


T* 


8P 


Eh 
O 


5 


© 
-P 


5 


© 
-P 




© 
-P 


© 
-P 


x 


bO 


-P 


-P 


-p 


-P 




-P 


-P 


c8 

> 


c8 


'S 


'a 


'a 


'£ 




'a 


1 


ft 


o 


o 


o 


O 




o 


o 


3 


c8 

108 


g 

■8 






a 

is 




.2 
"■p 




-p 
© 

I 


■a 

< 


1 

CD 


< 


c8 

-a 


O *-H 


c8 

a 


a 
§ 


8 c, 

sj 






bO 



© 
02 



ic8 

ft 



c3 ^ 

a 1 

0Q Q 



I 

o 



o 

< 



o 



60 





w 






8 


02 ^3 
© 3 




*a 


So 
o 


a -a 

$5 © 




© 


o 


© •? 




"a 


o 


a 




o 


.2 

*-p 


o 

<4H 






.°° 


*0 






13 


s 




T* 


o 






1 


1 


U <D 
O S 




I 


£ 


mo 
>> 




o 


o 


^2 






*5b 
o 


^ d 

© 5 


o 

u 

© 


'd 


13 


GO H 


^ 


© 
-P 


© 

© 


8 


-P 

'a 


O 


<l-dO 


o 



ffl 



© CO 
■P -P 

c8 Q. 


*0 

© 
-P 
-P 


© 
-P 
-P 


S S 


o 


o 






Q 



•d 

S 
-p 

'§ 
o 



Q 



ti © 03 



bO O 



a 

o 



O 









w 



© .£» 



J3*© 




3 & 


o 


frT* 




foA 


J 


© 


O 


ted 

opped 

ntheon 


•d 
© 
.2 




1 






l-H CQ 


© 
© 


Other 
substi 
when 
from 


o 
-p 

GO 

s 



TJ TJ T! 
© © © 



ass 

o o o 



W 5 W 



T5 



.a 

08 



o 



© 



o 



I 1 

o 



-a 
o 
O 



O 



ti 
H 



1 
.p 

a ,© © 

o o o 



o 

I 3 



h a 



61 



Indo-European Cattle-raiding Myth 

account of the content of the myth as we have reconstructed it, 
and the true task of interpretation is to account for the given data. 

As we have seen, the Indo-European *Trito myth contains 
two major elements. It is simultaneously a myth of the slaying 
of a monster and a myth of the first cattle raid. With regard to 
the former theme, I have tried to emphasize the explicit identi- 
fication of the monster as an outsider, a non-Indo-European, 
a thief and a usurper. Moreover, his serpentine form marks him 
as being in close connection with the earth. He is the aborigine, 
uncivilized and bound to his land, who opposes the I-E invader 
and meets defeat at his hands. His three heads may be yet 
another way of marking his foreign status, for, as Willibald 
Kirfel has demonstrated in his work of sweeping scope, Die 
Dreikopfige Gottheit, 67 the three-headed god is a major figure in 
the pantheon of pre-Indo-European peoples in India and the 
Mediterranean but never figured in that of the Indo-Europeans 
themselves. The description of the tricephaFs defeat is thus the 
description of the Indo-European victory. 68 

There is, moreover, an initiatory significance to this myth, as in 
so many other myths of monster slaying, 69 for the initiation of 
Indo-European warriors seems to have often involved a combat 
with a mock monster of triple form, as Dumezil argued early in 
his career. 70 In conquering this monster, the young warrior re- 
peats the events of primordial times which are related in our 
myth. He becomes again *Trito, the first I-E warrior, and he 
assimilates himself to the entire I-E onslaught that overthrew 
aboriginal opponents in every corner. 

The cattle -raiding aspect of the myth may be understood in 
similar ways. In order to do so, we must again note the enormous 
importance of cattle to the Indo-Europeans. Cattle were the very 
basis of the I-E economy, forming the essential measure of wealth 
and means of exchange. The animals supplied milk and meat, the 
main elements of the food supply; hides for clothing, blankets, 
bags, and shields; bones for tools; dung for fuel; and urine for 



67 Willibald Kirfel, Die Dreikopfige Gottheit (Bonn: Ferdinand Dummler, 1948), 
pp. 35-37, 83-84, 181-86. 

68 Ibid. 

69 On heroic initiation in general, see Mircea Eliade, Rites and Symbols of 
Initiation (New York: Harper & Row, 1958), pp. 81-87. Note also the important 
example of the Ngakola cult among the Mandja and the Banda cited in his 
Myths, Dreams and Mysteries (New York: Harper & Row, 1960), pp. 204-5. 

70 Georges Dumezil, Mythes et dieux des Germains (Paris: Ernest LeRoux, 
1939), pp. 92-106. 

62 



History of Religions 

disinfectant. 71 They were also crucial to the social order, serving 
as bridewealth and wergild. Society itself was understood as the 
collectivity not of only men, but of "men and cattle" or "bipeds 
and quadrupeds." 72 

Given this enormous importance of cattle within the socio- 
economic order, it comes as no surprise that the Indo -Europeans 
were always interested in preserving the cattle they had and in 
procuring more. Their prayers are filled with requests for cattle, 73 
and cattle raiding seems to have been one of the most important 
pursuits of the warriors. Certainly their heroic literature is filled 
with stories of such raids. To name just a few, one could cite the 
Irish epic Tain Bo Cualnge, the "Cattle Raid of Cooley," and the 
eleven other Irish stories which bear the title Tain Bo; 7 * Nestor's 
cattle raid (Iliad 11. 669-761) and the theft of Helios's cattle 
(Odyssey 12. 339-96) ; 75 the Armenian St. George's annual cattle 
theft; 76 and the innumerable cattle raids mentioned in the Rg 
Veda (e.g., RV 1.10.6-10; 3.16; 3.31.4-13; 3.53.9-14; 6.60.1-3; 
even the famous "Battle of the Ten Kings" is seen to be a cattle 
raid in RV 7.83.1). 

The *Trito myth tells of one such raid: more important, the 
first such raid. As such, it served as a model or prototype for all 
subsequent cattle raids. The mythic hero *Trito established the 
proper form of the raid for his I-E descendants, and among 
the Indo-Iranians he established the rites which insure its success: 
invocation of the warrior god, pouring of libations, and drinking 

71 On the importance of cattle, see Otto Schrader, Prehistoric Antiquities of 
the Aryan Peoples, trans. F. B. Jevons (London: Charles Griffin, 1890), pp. 259, 
284-87, 298, and Die Indogermanen, ed. H. Krahe (Leipzig: Quelle & Meyer, 
1935), pp. 23-29; Peter von Bradke, Vber Methode und Ergebnisse der arischen 
Alterthumswissenschaft (Giessen: J. Ricker, 1890), pp. 163-64; V. Gordon Childe, 
The Aryans (New York: Alfred A. Knopf, 1926), p. 82; Marija Gimbutas, "Proto- 
Indo-European Culture: The Kurgan Culture during the 5th, 4th and 3rd Millenia 
B.C.," in Indo-European and Indo -Europeans, ed. George Cardona et al. (Phila- 
delphia: University of Pennsylvania Press, 1970), pp. 157, 161, 190. 

72 J. Wackernagel, "Indoiranica: 9. Zum Dualdvandva," Zeitschrift fur 
vergleichende Sprachforschungen 43 (1910): 295-98; E. Benveniste, "Sur quelques 
dvandvas avestiques," Bulletin of the School of Oriental Studies 8 (1935-37): 
405-6. 

73 See, inter alia, RV 1.34.15, 3.31.20, 6.35.5; Yasna 4.5, 24.10, 62.10; Iguvine 
tables Via 30, VIb 13, Vila 30. On the Greek Hekatomb as a sacrifice which ob- 
tains 100 cattle rather than one in which 100 cattle are offered, see Paul Thieme, 
Studien zur indogermanischen Wortkunde und Religionsgeschichte (Berlin: Akademie 
Verlag, 1952), pp. 62-76; and also note Jaan Puhvel, "The Meaning of Greek 
j3oi»/cario?," Zeitschrift fur vergleichende Sprachforschung 79 (1964): 7-10. 

74 Weisweiler, pp. 27-29. 

75 On Greek cattle raiding, see Norman O. Brown, Hermes the Thief (New 
York: Vintage Books, 1947), pp. 3-7. 

76 Franz Cumont, "St. George and Mithra the Cattle Thief," Journal of Roman 
Studies 27 (1937): 63-71. 

63 



Indo-European Cattle-raiding Myth 

of intoxicants. In India, numerous texts show warriors identifying 
themselves with Trita (RV 2.11.19, 9.86.20, 2.34.14, 5.86.1), and 
the same is true for Iran with regard to Thraetaona (Yt. 19.92- 
96) 7T and Thrita, as can be seen from the name of a general of 
Xerxes recorded in Herodotus 7. 82: rpiTavralxy^is "Bold (Av. 
taxma-) as Thrita." 78 In this regard, the Roman emperor Com- 
modus's identification of himself with Hercules might not be 
taken as a sign of insanity, as it usually is, but merely as a re- 
emergence of the ancient I-E ideology. 79 

Finally, an ethical concern seems to be present in our myth, for 
it must be noted that *Trito's raid was not unprovoked aggression, 
but followed upon the tricephal's earlier theft. It is thus justified, 
for the I-E hero is only taking back that which rightfully belongs 
to his people. Moreover, he uses open force to regain his stock, in 
contrast to what must have been regarded as the despicable stealth 
of the tricephal. The myth is an imperialistic myth, it is true, but 
even imperialists need their rationalizations. 80 

As I stated at the outset, I feel that this myth of *Trito, 
"Third" is to be grouped with that of *Manu and *Yemo, "Man" 
and "Twin." The Indo-Iranian figures are often grouped together, 81 
and in Livy the story of Hercules and Cacus (1. 7. 3-12) follows 
immediately upon that of Romulus and Remus (1. 6. 3-7. 2) — 
the figures who represent *Manu and *Yemo — for no reason that 
is readily apparent from the immediate context. 

While the myth of "Man" and "Twin" seemed to be a myth of 
the sovereign function, establishing the model for later priests 
and kings, so this myth of "Third" seems to be a myth of the 
warrior function, establishing the model for all later men of arms. 
As yet, I have not been able to locate a myth which relates in a 
similar fashion to Dumezil's "Third Function," but that is not 
to say that one does not exist. Perhaps it will be discovered at some 

77 See the treatment of this text in Gerhard Binder, Die Aussetzung des Konigs- 
kindes Kyros and Romulus (Meisenheim am Glan: Anton Hall, 1964), pp. 60-61. 

78 Herman Guntert, Der arische Weltkonig und Heiland (Halle: Max Niemeyer, 
1923), p. 29, was the first to recognize the significance and importance of this 
name. 

79 For another creative use of this old image, see the analysis of Pindar's first 
Pythian Ode in Jurgen Trumpf, "Stadtgrundung and Drachenkampf," Hermes 
86 (1958): 129-57. Again it is a case of a monarch attempting to bolster his reign 
by identification with the figure of Herakles. 

80 Rose's argument that the aborigines told this myth to explain their loss of 
cattle to the I-E's (see n. 61 above) need not be dismissed on this account, for 
such a myth could be told by both sides in the conflict. This is the case with the 
cattle -raiding myth of the Nuer and Dinka (see E. E. Evans-Pritchard, The 
Nuer [Oxford: Oxford University Press, 1940], pp. 125-26). 

81 RV 10.64.3, 1.163,2-3, 8.52.1; Yasna 9.3-13, Yast 19.36. 

64 



History of Religions 

later date, or it may simply be that this ''function," being more 
disparate and ill-defined than the others as well as lower in rank 
(I prefer the term "commoners" to his agriculteurs or eleveurs), 
may simply lack such a mythical prototype. Only further re- 
search will determine the case. 

University of Chicago 



65